United States Patent [191 ONeal, Jr.
[451 Aug. 27, 1974 WELDING CONTROL APPARATUS [75] Inventor: George ONeal, Jr., Plymouth,
Mich.
[73] Assignee: Weltronic Company, Southfield,
Mich.
[22] Filed: July 24, 1970 [21] Appl. No.: 58,126
[52] U.S. Cl. 219/110 [51] Int. Cl B23k 9/10 [58] Field of Search 2l9/l081l0,
219/114; 323/22, 24 V, 24 SC, 102; 328/70, 71, 72, 78, 81, 84
[56] References Cited UNITED STATES PATENTS 2,234,963 3/1941 Coffin 219/114 X 2,866,134 12/1958 Hartwig 219/114 2,985,816 5/1961 Scholtes et a1 219/114 3,005,947 10/1961 Scholtes et al....
3,202,871 8/1965 Shelar 219/131 3,243,689 3/1966 Perrins 323/22 SC 3,452,283 6/1969 ONeal, Jr 219/114 X 3,486,042 12/1969 Waltrous 323/22 SC FOREIGN PATENTS OR APPLICATIONS 637,644 5/1950 Great Britain 323/102 Primary ExaminerJ. V. Truhe Assistant ExaminerClifford C. Shaw Attorney, Agent, or FirmHamess, Dickey & Pierce [57] ABSTRACT A firing control System for controlling controllable rectifier devices to control the transfer of energy from a source of energy to a workpiece to be welded and for establishing controlled firing sequences and times.
119 Claims, 11 Drawing Figures I PAIENTEDMIBZTW 1 a:m.-518f smears 72 2%?" hind! i/l l INVENTOR.
2$? LZ AQ@L AZ W 9" PAlENTimusznsn v 331%.518 y I smears INVENTOR.
Y ll/I25.
WELDING CONTROL APPARATUS BACKGROUND AND SUMMARY OF THE DISCLOSURE This invention relates to control apparatus and more particularly to a firing system suitable for use in conjunction with ignitrons or similar controlled rectifier devices to control transfer of energy from an energy source to a workpiece to be welded.
In general, the system preferably employs, in substantial part, solid-state devices and certain of the features of the invention are directed to the solution of the problems that arise as a result of the effort to obtain the known advantages of solid-state devices.
An object of this invention is to improve solid-state firing systems for ignitrons or the like.
Another object of this invention is to improve separate-excitation firing circuits.
Another object of the invention is to prevent line voltage transients from adversely affecting the operation of a firing system including solid-state devices.
Another object of this invention is to increase the safety of operation of firing circuits for ignitrons and the like.
Another object of this invention is to disable a firing circuit to fire an ignitron or the like in a welding system until a preselected period has elapsed following the receipt of an initiating signal.
Another object of the invention is to improve circuits for establishing the proper sequence and alternation of operation of a pair of back-to-back connected contactor devices.
Another object of the invention is to establish in a precise and selectable interval between the instant of initiation of a weld interval and the time at which welding current can first be delivered to the workpiece.
Another object of this invention is to improve the accuracy of timing a heat control apparatus.
Another object of the invention is to increase the effective range of control of a heat control apparatus.
A further object of the invention is to insure proper operation of a separate-excitation firing system when utilized in conjunction with a resistance welder operating with an inductive load.
Another object of the invention is to prevent the separate-excitation firing system from prematurely firing control ignitrons or the like.
Another object of the invention is to disable a separateexcitation system from energizing the input circuit of an ignitron or the like until the anode voltage of that ignitron has appropriately changed.
A further object of this invention is to automatically adjust for power factor in a resistance welding system.
The manner in accomplishing the foregoing objects and other objects and features of this invention may be perceived from the following detailed description of an embodiment of the invention when read with reference to the accompanying drawings, in which:
FIG. 1 is a schematic representation of a portion of an electrical control circuit embodying certain of the principles of the present invention;
FIG. 2 is a schematic representation of another portion of the circuit of FIG. 1 and should be placed below FIG. 1 for proper orientation;
FIG. 3 is a schematic representation of a portion of an electrical control circuit embodying certain of the principles of the present invention;
FIG. 4 is a schematic representation of another portion of the electrical control circuit of FIG. 3;
FIG. 5 is a graphical representation of certain electrical relationships which can exist in the circuit of FIGS. 3 and 4;
FIG. 6 is a graph illustrating the phase relationship of the line voltage and current and the voltage versus time relationship of the timing pulse for the system of the present invention;
FIG. 7 is a schematic representation of a portion of another electrical control circuit embodying other principles of the present invention;
FIG. 8 is a schematic representation of still another portion of the circuit of FIG. 7;
FIG. 9 is a schematic representation of another portion of the electrical control circuit of FIG. 7;
FIG. 10 is a schematic representation of another portion of the electrical control circuit of FIG. 7; and
FIG. 11 is a schematic representation of another portion of the electrical control circuit of FIG. 7.
For convenience of illustration, the transformer windings have been illustrated in the drawings in a way to best illustrate the functions of those transformers and consequently the primary and secondary windings are shown separated. Common prefix designations have been employed in each case, however, to permit identification of which secondary windings areassociated with which primary windings. Additionally, in the drawings, the sources of direct voltage have been indicated by a circle bearing a sign indicative of the polarity of the source. It is to be understood that in each case the other terminal of the source is assumed to be connected to ground. For convenience and clarity, voltage values have been referred to in the following description. It is to be understood that they are but representative.
In general, the circuits illustrated on FIGS. 1 and 2 of the drawings comprise a pair of ignitrons [G1 and [G2 (or other controlled contactor means) for selectively connecting a source of energy S1 to a welding transformer WT for controlling the application of energy to a workpiece WP which is to be welded. The ignitrons IG] and IG2 are controlled by individual firing circuits including controlled rectifiers lCRE and 2CRE. Those firing circuits are operated under the control of driving circuits including transistors Q17 and Q18, and transistors Q19 and Q20. Those driving circuits are, in turn, controlled by four separate circuits including a delayed firing system, comprising unijunction transistor IUJ and transistors Q1 through Q4, an ignitron anode voltage sensing system comprising transistors Q5 through Q7, a lead-trail control circuit comprising transistors Q8 and Q9, and a heat control circuit comprising transistors Q10 through Q16.
In the customary present commercial practice, ignitrons, or the like, are customarily employed with socalled anode firing circuits in which the voltage applied across the ignitron also serves as the energizing or plate voltage for the controlling or firing device, such as a thyratron, in the firing circuit. In such systems, the peak line voltage, which can be high due to transients, is in large part applied directly across the firing device which has made it difficult satisfactorily to adapt the system to the use of solid-state firing devices in view of their sensitivity to voltage transients. If the rate of voltage rise is sufficient, as it can well be with line-voltage transients, the firing device can, improperly, fire even though there be no input signal.
In the present arrangement, the problems arising from the transient-voltage sensitivity characteristics of solid-state devices, such as silicon controlled rectifiers, are effectively solved by isolating the firing devices from the line. Specifically, a separateexcitation circuit is employed and filtering means are employed between the source of voltage and the device to suppress transient voltage peaks. As a further means, additional filtering means are or may be provided between that device and the ignitron to control the rate of increase of the igniter current.
Among the other advantages which accrue from the disclosed arrangement is the fact that it provides a wide latitude in the selection of the firing angle of the ignitrons, and they can be fired earlier in the half cycle than in the customary commercial circuits.
The two ignitrons 1G1 and [G2 are connected in back-to-back or antiparallel relationship between the source S1 and the primary winding of the welding transformer WT, in a manner well known in the art. The operation of these ignitrons is controlled by firing circuits including devices lCRE and 2CRE. To effectively preclude dangerous premature firing of the ignitrons, switch NWCR is actuated to indicate that the apparatus is prepared for welding. In a common practice, welder control circuits include a timer having a relay, often referred to as the no-weld relay," which is actuated upon initiation of the squeeze interval provided the apparatus is otherwise in condition to weld, and it is contemplated that the switch NWCR illustrated in FIG. 1 of the drawings may, and normally will be, a contact of that or of a counterpart relay. The no-weldcontrol relay may well not operate at a point of zero line voltage and could result in a transient signal which would improperly actuate the firing circuit. The illustrated circuits obviate this possible malfunctioning.
When switch NWCR is closed, a circuit is completed from the source S1, through that switch and through fuses F1 and F2 to energize the transformer primary winding TIP, which is inductively coupled to secondary windings T151 and T1S2 of that transformer. Secondary windings T181 and TlS2 are connected in outof-phase relationship and the phase relationships of the several windings of that transformer are indicated by the dot placed adjacent one end of each of the windings TlP, TlSl and T1S2 to denote those winding ends which are of the same polarity at a given instant.
When the voltage across winding TlP is such that the left-hand end of that winding is positive relative to the right-hand end, for example, the voltage induced across secondary winding TlS2 is such that its upper end is positive relative to its lower end. Under that condition, current flows in a circuit including resistor RSlb and rectifier 23RE to charge capacitor 21C so that its upper electrode becomes positive relative to its lower electrode. As will be seen, the energy stored by capacitor 21C is utilized to fire ignitron 1G2. Charging resistor RSlb may be provided as a separate element, but in a constructed embodiment of the invention, the effective resistance of the secondary winding T152 was found to be adequate and resistor RSlb is illustrated in dotted lines to connote that it represents the internal resistance of the winding.
It will be observed that during this same half cycle, the polarity of the voltage across secondary winding TlSl is such that rectifier 22RE blocks current flow and hence capacitor 20C does not charge during this half cycle. However, during the subsequent half cycle, in which the polarity is reversed, capacitor 20C is charged in a manner similar to that above described in connection with capacitor 21C in preparation for the firing of ignitron 101.
Thus, at the end of one full cycle of the current from source S1 following the operation of switch NWCR, both capacitors 20C and 21C are charged in preparation for the operation of the system. Until they charge, the firing circuits cannot actuate the ignitrons. This one-cycle delay serves as a safety measure to insure that the welding electrodes will have engaged the workpiece before welding current is applied to the welding transformer. In the customary timers, squeeze time must be initiated before switch NWCR will close and the necessity of charging capacitors 20C and 21C in order to enable the firing circuits to fire the ignitrons [G1 and 102 requires that the squeeze time be at least one cycle in duration.
In the preferred arrangement, capacitors 20C and 21C are made sufficiently large to store a substantial amount of energy which may be abruptly discharged through the input circuits of the ignitrons. Once those capacitors are charged, this energy is available for application to those circuits, but cannot be so applied until the control rectifier devices lCRE and 2CRE have a suitable gating potential applied to their control electrodes or gates.
Means are provided for applying gating pulses to the silicon controlled rectifier devices lCRE and 2CRE in selectable timed relation to the voltage applied to the anodes of the ignitrons. When the gating signal is applied to controlled rectifier 2CRE (during the halfcycle of the source voltage in which the anode of ignitron IG2 is positive relative to its cathode), that recti fier is rendered conductive to establish a discharging path for capacitor 21C through that rectifier, inductor or choke CH2, resistor 87R, fuse F5, through'the igniter-cathode path in tube 102 and back to capacitor 21C. In the preferred arrangement, capacitors 20C and 21C are preferably of substantial capacitance (such as 20 microfarads) so that a substantial amount of energy can be delivered to the input circuits of the ignitrons. It is a characteristic of the preferred controlled rectifiers lCRE and 2CRE that when conduction is initiated by virtue of the application of an input signal to their gates, the termination of the gating pulse will not in and of itself terminate conductivity of those devices.
In response to the signal applied to the input circuit of ignitron 102, that ignitron will conduct between its anode and cathode, producing energization of the welding transformer WT from the source S1 and a resultant application of a pulse of energy to the workpiece WP. In a similar manner, ignitron IGl is rendered conductive at a selectable point in that half-cycle dur-' ing which its anode is positive relative to its cathode to similarly energize transformer WT to apply a pulse of energy to the workpiece WP.
It will be observed that the firing energy storage means, such as a capacitor 21C, together with the resistance in its charging circuit, such as the illustrated internal resistance RSlb of transformer winding T182, constitute a resistance-capacitance low-pass filter or integrating network. As a result, if the line voltage abruptly and transiently changes, tending to induce a transient voltage peak across secondary winding T1S2, that resistance-capacitance network will effectively suppress the voltage peak from appearing at the controlled rectifier device 2CRE and effectively preclude any such transient from producing a sufficient rate of change of current to cause that device improperly to become conductive.
It will further be noted that the circuit including choke CH2 and the resistance in the discharging circuit for capacitor 21C, including resistor 87R, constitute a filter for limiting the rate of change of the discharge current of capacitor 21C so as to limit the magnitude of the current in the input circuit of the ignitron.
Resistors 78R and 85R, which are connected in parallel with capacitors 21C and 20C, respectively, are preferably of sufficiently large resistance so that they do not significantly affect the nonnal operation of the circuit. In a practical embodiment, those two resistors were selected to have a value of about 50,000 ohms so that the network including the associated capacitor had a time constant of one second, which is large relative to the normal interval between the charging of the capacitor and the time at which firing circuit will be triggered to apply the pulse of energy to the ignitron. However, at the termination of the operation, when switch NWCR is opened, those resistors serve to discharge their associatedcapacitor as a safety measure.
The gate signals for the controlled rectifier devices lCRE and 2CRE are applied through pulse transfonners T3 and T4, respectively. Any pulse appearing across the secondary winding of transformer T4, for example, which is of a polarity such that the left-hand end of that winding is positive relative to the right-hand end, is dissipated through rectifier 20RE and resistor 74R. A pulse of the opposite polarity is applied to the input or gate circuit of controlled rectifier 2CRE via resistor 74R to cause that device to apply adischarge pulse from capacitor 21C to the input circuit of ignitron [(32, as above described. The network comprising capacitor 23C and resistors 74R and 75R are elements of a circuit for filtering high-frequency spurious transients and for effectively preventing improper actuation of the controlled rectifier device by transients. The gating pulses applied through transformer T4 are of sufficient magnitude to produce gating of the device despite this filtering or desensitizing network. The other firing circuit operates in a similar manner.
The application of pulses to the pulse transformers T3 and T4 is controlled by the driving circuit comprising transisters Q17 and Q18 and the driving circuit comprising transistors Q19 and Q20, respectively. These driving circuits are controlled by a delayed firing system via conductor 10, by an ignitron anode voltage sensing system via a conductor 12, by a lead-trail control circuit which is connected to the two driving circuits via leads 14 and 16, respectively, and by a heat control circuit via conductor 17. As will be seen, in the illustrated arrangement, eachof these leads may be at either of two selected voltages. ln the illustrated arrangement these have been selected to be a positive voltage (such as positive 12 volts) and ground.
Conductor is connected to the bases of transistors Q17 and Q19 through resistors 48R and 54R, respectively; conductor 12 is connected to the bases of those transistors through resistors 49R and 55R, respectively;
conductor 14 is connected to the base of transistor Q17 through resistor 52R; conductor 16 is connected to the base of transistor Q19 through .resistor 53R; and conductor 17 is connected to the bases of transistors Q17 and Q19 via resistors 50R and 51R, respectively.
The emitters of transistors Q17 and Q19 are grounded and their collectors are connected to a source of positive potential through load resistors 57R and 56R, respectively. Negative biasing voltages are applied to the bases of those transistors through resistors 78R and 79R, respectively. If any one of the conductors 10, 12, 14 or 17 is at the noted positive potential (assumed to be 12 volts), transistor Q17 is biased effectively to saturation, and similarly, if any one of the conductors 10, 12, 16 or 17 is at the noted positive potential, transistor Q19 is biased effectively to saturation. These input networks therefore constitute, in effect, or gates under which if any one of the noted conductors associated with transistors Q17 or Q19 is at its positive potential or state, the associated transistor is biased effectively to saturation. Under that condition, the collector of that transistor is at a relatively low potential, herein assumed to be ground potential. However, at the instant that all of the noted conductors associated with the transistor concurrently reach the lower (ground) potential, the voltage at the base of that transistor drops sufficiently to render that transistor effectively non-conductive. For example, whenever all of the conductors 10, 12, 16 and 17 concurrently reach ground potential, transistor Q19 is rendered nonconductive and as a result a positivegoing pulse is applied through the capacitor 14C to the base of transistor Q20. The emitter of transistor Q20 is grounded, and the collector is connected to a source of positive potential through the primary winding of transformer T4 and via switch SW1. The base is connected to a source of negative potential through resistor 59R. As a result of the application of the positive pulse to the base, transistor Q20 conducts current from the positive source through switch SW1, and through the primary winding of transformer T4 so that a pulse is induced in the secondary winding of that transformer. The shape and duration of the pulse which is applied to the base of transistor Q20, and hence the shape and duration of the pulse applied to the controlled rectifier device 2CRE via transformer T4 is controlled by means including resistor 56R, capacitor 14C and the resistance of the base of transistor Q20. Rectifier 4RE serves to prevent any substantial negative voltage from being applied to the base of transistor Q20. Rectifier 19RE serves to dissipate the voltage which is induced across the primary winding of transformer T4 upon the collapse of the magnetic field at the termination of conduction of transistor Q20 at the end of the pulse.
The driving circuit including transistors Q17 and Q18 operates in a similar fashion, producing a pulse of energy at transformer T3 in the event that and when the voltages on conductors 10, 12, 14 and 17 all reach their lower or ground potential. It will be noted that switch SW1 also controls the application of positive voltage to the collector of transistor Q18. This switch is provided as a further safety measure and preferably is a contact of or is controlled by the weld-no-weld switch customarily provided in resistance-welder timers and which must be closed in order for welding to proceed. Whenever that switch is open, the driving circuits are incapable of applying pulses through transformers T3 and T4 to the firing circuits.
The lead-trail circuit comprising transistors Q8 and Q9, (FIG. 2) controls, via conductors l4 and 16, which of the two driving circuits and hence which of the two firing circuits can operate at any time, and alternately enables those circuits. This circuit is energized via a transformer, the primary winding TlOP of which is illustrated to be connected across the source S1 and the secondary winding TlOS of which appears on FIG. 2. When the upper terminal of the secondary winding TlS is positive with respect to the grounded center tap, which occurs when the left-hand terminals of source S1 and primary winding TP are positive relative to their other terminals, current flows through rectifier 9RE, resistor 66R, resistor 68R, and via the base and emitter of transistor Q8 back to the grounded cenw ter tap of secondary winding T108. The base of transistor Q8 is connected to a source of negative potential through resistor 34R, the emitter is grounded, and the collector is connected to a source of positive potential pthrough load resistor 10R. When the base is driven positive, as described, transistor Q8 conducts substantially at saturation and the voltage at its collector drops effectively to ground potential. This voltage is applied via conductor 14 and through resistor 52R to the base of transistor Q17 in the driving circuit associated with ignitron 1G1. This is an enabling signal, which, other conditions met, will permit the firing circuit associated with ignitron 1G1 to fire that ignitron, and it will be observed that this occurs during the half cycle in which the anode of ignitron IGl is positive with respect to its cathode.
During the same half-cycle, the lower terminal of transformer secondary THIS is negative with respect to ground so that transistor O9 is effectively nonconductive in view of the connection of its base to a source of negative potential through resistor 35R. Consequently, a positive voltage (e.g., 12 volts) is applied via conductor 16 and through resistor 53R to the base of transistor 019 to disable that driving circuit and the firing circuit including controlled rectifier 2CRE to tire ignitron 1G2.
In the preferred arrangement, the turns ratio of the transformer including windings TlOP and T10S is selected so that a fairly high secondary voltage is applied to the lead-trail circuit such as, for example 1 15 volts. As a result, a quite substantial peak positive voltage is applied through rectifier 9RE and resistor 66R. However, rectifier 16RE is connected between a point at the junction of resistors 66R and 68R to a source of much lower positive potential (e.g., 12 volts). Accordingly, rectifier 16RE tends to clamp the voltage at the junction between resistor 66R and 68R to insure that the maximum positive voltage applied to the base of transistor Q8 does not exceed that selected value. Since transistor Q8 will operate with a voltage applied to the base of substantially less than that value, the operating point for that transistor occurs early in the half cycle of the line voltage, approaching quite closely the zero degree point. Transistor Q8 will remain conductive throughout essentially the complete half-cycle and during the alternate half-cycle will be non-conductive, whereas transistor Q9 will be conductive in the opposite half-cycle. Rectifiers 24RE and 13RE are provided to prevent negative pulses from being applied to the bases of transistors Q8 and Q9, respectively.
The lead-trail circuit, when connected as illustrated, operates continuously whenever the source S] is connected, operating transistors Q8 and Q9 during alternate half cycles so as alternately to enable the firing circuits associated with ignitrons 1G1 and IG2. In order to control when welding occurs and additionally to provide a means for effectively preventing firing during the first half-cycle thereafter so as to prevent saturation of certain types of welding transformer cores, the delayed firing system illustrated in FIG. 1 of the drawings, is provided. The delayed firing system, including uninjunction transistor IUJ and transistors Q1 through O4, is actuated by a weld signal applied to conductor 20. An appropriate signal is conventionally available in timers associated with present commercial welding equipment. This signal normally is applied when the timer has been set and desirably is synchronized with the voltage from source S1 so that the operating signal is both applied and removed at the zero degree points of the source voltage. In the illustrated arrangement, it is assumed that the weld signal applied to conductor 20 is at an appropriate positive value (e.g., 12 volts) and that the conductor 20 is grounded in the absence of a weld signal.
When the positive weld signal is applied to conductor 20, it appears across a network comprising resistor 30R, variable resistor VRl and capacitor 1C. The voltage across capacitor 1C is applied to the emitter electrode 21 of unijunction transistor IUJ, one base electrode 22 of which is connected to a source positive potential through resistor 1R. The other base electrode 24 of that double-base unijunction transistor device is connected to ground through a resistor 4R. Capacitor 1C charges at a rate controlled by resistor 30R and variable resistor VRl, which selects the delay established by the delayed firing system. When the voltage applied to emitter 21 rises to a sufficient value, the impedance of the unijunction device IUJ between the emitter and base electrode 24 abruptly falls and capacitor 1C discharges over a path including electrodes 21 and 24 and resistor 4R. When capacitor IC has discharged sufficiently, the voltage applied to emitter 21 falls below the voltage required to maintain the forward bias condition and unijunction device IUJ changes to a high impedance condition so that capacitor 1C can again commence to charge via resistor 30R and variable resistor VRl. This operation repeats, in the nature of a relaxation oscillator, producing a series of positive-going short-duration pulses across resistor 4R. The time between the receipt of the weld signal via conductor 20 and the first such pulse is closely established, but the oscillatory rate is not synchronized or necessarily related to the frequency of the source S1.
The positive-going pulse appearing across resistor 4R is applied via resistor 12R to the base of transistor Q1. The base of transistor O1 is connected to the source of negative potential through resistor 38R, the emitter thereof is grounded and the collector is connected to the source of positive potential through load resistor 5R. Transistor O1 is driven effectively to saturation and as a result a relatively large amplitude negative-going pulse is applied through capacitor 3C and rectifier IRE to the base of transistor Q2. Transistors Q2 and Q3 are cross-coupled to form a flip-flop circuit, with the collector of transistor Q2 being coupled to the base of transistor Q3 by the network comprising capacitor 5C and resistors 26R and 33R and with the collector of transistor Q3 being coupled to the base of transistor Q2 through a similar network compressing resistors 25R and 32R and capacitor 4C. The collectors of transistors Q2 and Q3 are connected to a source of positive potential through load resistors 6R and 7R, respectively, the bases of those two transistors are connected to a source of negative potential through resistors 32R and 33R, respectively, and the emitters of those transistors are grounded. Transistors Q2 and Q3 conduct alternatively and desirably means are provided for insuring that prior to the receipt of the described pulse, transistor Q2 is conducting effectively to saturation whereas transistor Q3 is cut off. In the illustrated circuit, a signal derived from the timer is employed to serve this function. This signal applied to conductor 26, is assumed to be positive voltage (e.g., 12 volts) which is applied to conductor 26 at all times before operation of the timer is initiated. This signal could, of course, be derived from any other suitable source.
The positive voltage on conductor 26 is applied to the base of transistor Q2 through resistor 31R and serves to maintain transistor Q2 in a conductive state. When that signal terminates at the time of initiation of the timer (and prior to the application of a signal to conductor 20), the circuit comprising transistors Q2 and Q3 remain in the same state, with transistor Q2 conducting and transistor Q3 non-conductive. However, when the abrupt negative-going signal is applied to the base of transistor Q2 via rectifier IRE, transistor O2 is cut off, its collector voltage abruptly rises, and the resulting positive-going signal is applied via capacitor C and resistor 26R to the base of transistor O3 to render that device conductive. The circuit regeneratively switches its state. The successive negative pulses applied to the base of transistor Q2 via rectifier IRE from transistor Rl will not be effective to change the state of transistors Q2 and Q3 since transistor O2 is already non-conductive under this condition.
When transistor O2 is rendered non-conductive, the collector voltage rises abruptly and this signal is applied to the base of transistor Q4 via resistor 40R, that base being connected to a source of negative potential through resistor 46R. The collector of the grounded emitter transistor Q4 is connected to a source of positive potential through load resistor 8R, and when transistor Q4 conducts in response to the positive signal applied to its base, its collector abruptly drops from the iniial positive potential (e. g, 12 volts) to a lower potential, such as ground, and this signal is applied through rectifier 4RE to the conductor which is connected to the driving circuits as previously described. It will be recalled that when this signal is applied to conductor 10, the driving circuits (including transistors Q17 through Q20) are enabled to operate as far as this particular control is concerned, that is, the delayed firing system is no longer able to prevent operation of the ignitrons by the firing circuits.
It will be observed that when transistor O3 is conducting, its base, and hence the upper electrode of capacitor 6C is at a potential substantially below the voltage applied to the lower electrode of that capacitor via conductor 20, that is, lower than the weld signal voltage. As a result, when the weld signal terminates so that conductor 20 becomes abruptly grounded (which desirably occurs at the zero degree point of the sine wave from source S1), a negative pulse is applied from capacitor 6C through rectifier ZRE to the base of transistor O3 to render that transistor non-conductive and to thereby cause transistor Q2 to become conductive. Since the termination of the weld signal applied to conductor 20 also terminates the operation of the oscillatory circuit including capacitor 1C and unijunction de vice 1U) there will be no additional pulses applied to the base of transistor Q2 so that the circuit comprising transistors Q2 and Q3 will remain in this state until the next weld signal is received.
It will be seen that the delayed firing system applies a disabling voltage to conductor 10 at all times that the weld signal is not being applied to conductor 18, but that it will shift the potential on conductor 10 to an enabling voltage a timed interval after the weld signal is applied to conductor 18. Since the weld signal appears at the beginning of the weld interval, the delayed firing circuit will be effective during the first half-cycle of the line frequency, only, to prevent firing of either of the ignitrons for a preselected interval even though other elements of the circuits may indicate that welding may proceed. It is presently believed that the optimum delay period is 87 after the zero degree point of the sine wave of the source S1 at which the weld signal is applied to conductor 20. This interval may be varied to accommodate variations in the power factor of the load by variable resistor VRl. It is desirable, however, that the magnitude of the delay be quite precisely selectable, which is one of the reasons for the present preference for'the unijunction device lUJ for this function. The discharge point of such devices, in properly designed circuits, is substantially independent of supply voltage variations. Additionally, with proper circuit design including desirably the use of a temperature compensated resistor 1R, the emitter voltage at which the device will discharge is substantially independent of temperature. Further, since the device operates at an input voltage which is quite low relative to the voltage on conductor 20, the essentially linear portion of the charging curve of capacitor 1C is utilized. As a result the illustrated system provides extremely precise timing of the interval betweenthe application of an appropriate portntial to the weld line 20 (at the zero degree point) and the instant at which the firing circuits are enabled to fire during the first one-half cycle of operatron.
It should again be noted that this delayed firing system does not necessarily cause firing of the ignitrons but merely establishes a minimum firing angle for the first half-cycle, and that after the first half-cycle of any weld, it is ineffective to interfere with the free selection of the firing points of the ignitrons.
The heat control circuit, comprising transistors Q10 through Q16, selectively controls the firing angles of the ignitrons I01 and 162 to control the percent heat and hence the magnitude of the energy delivered to the workpiece WP. In general, the heat control has a capacity to produce firing of those ignitrons at any selected phase angle provided the other conditions established by the circuitry are met. Among those other conditions of course, in the illustrated arrangement, is that if the heat control be set to fire the ignitrons at a phase angle less than a selected value in the order of 87 /2, no such firing will occur during the first half-cycle of the weld until after the minimum delay angle which is established by the delayed firing system.
The alternating current signal appearing across the secondary winding T10S (FIG. 2) is synchronized with the source S1. This signal is full-wave rectified by rectifiers RE and llRE and applied through resistor 65R to one electrode of rectifier RE, the other electrode of which is connected to a source of negative potential. Rectifier 15RE prevents the voltage on conductor 32 from becoming more negative than a selected value, such as negative 12 volts. If the magnitude of the voltage of the negative peaks of the full-wave rectified sig nal be large relative to that selected negative 12 volt value, then the voltage on conductor 32 will be in the form of a negative 12 volt signal with a positive-going (to ground) spike each 180.
This signal is applied through resistor 72R at the base of transistor Q10, that base being connected to a suitable source of positive potential through resistor 82R. Transistors Q10 and Q11 are interconnected as a multivibrator in a form of Schmitt trigger circuit, with the collector of transistor Q10 being coupled to the base of transistor Qll via a network comprising capacitor 25C and resistors 44R and 22R, and with the emitters of the two transistors being coupled via resistor 63R. When the voltage on conductor 32 is at the negative 12 volt level, transistor Q10 is held in a non-conductive state and transistor Q11 is conducting. At the positive-going input signal applied via conductor 32 to the base of transistor Q10, transistor Q10 begins to become conductive and as a result of the coupling between transistors Q10 and Q11, transistor Q10 becomes fully conductive very rapidly and transistor Q11 is driven below cutoff. The magnitude of the input voltage to the base of transistor Q10 at which this triggering will occur is quite precise and repetitive and the point at which the triggering occurs in relation to the voltage of source S1 can be precisely selected by selection of the parameters of the trigger circuitry, by selection of the turns ratio of transformer T10 to control the magnitude of the ac. voltage across secondary winding TlOS, and by selection of the magnitude of the negative biasing voltage applied to rectifier 15RE. In a constructed arrangement, with 115 volts across the secondary winding T105, the circuit comprising transistors Q10 and Q11 was accurately triggered 10 in advance of the zero degree point (and the [80 point) on the ac. wave form, transistor T10 being rendered conductive and transistor Q11 being rendered non-conductive.
The trigger remains in this condition until the positive signal diminishes toward the selected negative 12 volt point and in the constructed embodiment, this occurred at about 10 after the zero degree point (and the 180 point) of the wave form of the source S1. At that time, transistor Q10 again becomes non-conductive and transistor Qll again becomes conductive. When transistor Q11 is conducting, its collector voltage is at a relatively low value, approaching ground. When transistor Q11 is non-conductive, at each pulse on conduc- I tor 32, its collector voltage is at a higher voltage such as 12 volts positive. Consequently, during the operation of the circuit, conductor 34 is supplied 120 times per second with a positive-going (from ground) to positive 12 volts) essentially square-wave pulse of relatively short (e.g., duration and having its leading edge accurately related to and in advance of (e.g., l0) of the zero degree point (and 180 point) on the ac. wave form of source S1.
The pulses are applied to the base of transistor Q12 by a network comprising capacitor 9C and resistors 61R and 45R. Transistor 012 is rendered conductive by each such pulse to apply a corresponding series of negative-going pulses to a multivibrator circuit comprising transistors Ql4 and Q15 via a network including capacitor 10C and rectifier SRE. The multivibrator including transistors Q14 and Q15 is similar to the multivibrator comprising transistors Q2 and Q3 (FIG.' 1) above described and operates in a similar fashion. Transistor Q14 is normally cut off and transistor Q15 is normally conducting.
At each of the short-duration negative-going pulses applied to the base of transistor Q14 by transistor Q12, transistor Q14 is turned off and transistor Q15 is turned on. When transistor Q15 is triggered to its conductive state, its collector voltage falls essentially to ground potential and this voltage is applied via conductor 15 and a network including resistors 43R and 47R to the base of transistor Q13 to block conduction in that transistor.
As a result, the collector voltage of transistor Q13, at
conductor 17, is approximately 12 volts positive. This signal is applied through resistor 50R to conductor 18 to disable the driving circuit comprising transistors Q17 and Q18 and is applied via resistor 51R to conductor 19 to disable the driving circuit comprising transistors Q19 and Q20. It will be noted that this occurs slightly (e.g., 10) before the cycle commences.
When transistor Q14 is rendered non-conductive just prior to the beginning of a cycle, the potential at its collector rises and is applied across the network comprising variable resistors VR2 and VR3, resistor 29R, and capacitor 2C. Capacitor 2C charges at a rate determined by the resistance of the charging circuit. Variable resistors VR2 and VR3 are provided to permit selection of the per cent heat and the power factor adjustment, respectively. Resistor 29R establishes the maximum heat for which the system can be set, with VR2 and YR3 set to their minimum resistance positions.
The charge on capacitor 2C is applied to the emitter of unijunction transistor 2UJ which functions in the same manner as unijunction device 1U], previously described. When the voltage across capacitor 2C rises to a sufficient value, unijunction device 2U] operates to apply a positive-going pulse to the base of transistor Q16 through a network comprising resistors 13R and 39R and capacitor 12C. Transistor Q16 is rendered conductive and desirably saturates, and its collector voltage drops from, say, 12 volts to approximately ground potential to develop a negative-going pulse which is applied through capacitor 11C and rectifier 6RE to the base of transistor Q15 to restore the trigger circuit comprising the transistors Q14, Q15 back to its original state. The reestablishment of conduction in transistor Q14 effectively removes the charging source for capacitor 2C. The termination of conduction through transistor Q15 results in the application of ground potential via conductor 15 and resistor 43R to the base of transistor Q13 to cause that device to become fully conductive. As a result, its collector voltage at conductor 17 drops substantially to ground potential which is communicated to conductors l8 and 19 through resistors 50R and 51R to enable the driving circuis and the firing circuits to fire the ignitrons 1G1 and 162, as far as this control is concerned. As will be seen, in the normal operation of the circuit, all of the other conditions enabling one of the two firing circuits to operate have normally been met prior to the receipt of this heat-control signal so that normally it is the application of ground potential to conductor 18 and 19 which actually produces the firing of the appropriate one of the two ignitrons 1G1 and 1G2.
The ignitron anode voltage sensing equipment including transistors Q through Q7 serves to overcome that which has been a serious disadvantage of separate excitation types of firing systems. The apparatus thus far described will function satisfactorily but is subject to possible misfiring with highly inductive loads. Thus, if the load current trails the load voltage due to the inductive reactance of the load, the ignitron which is fired during one-half cycle may continue to conduct even through the phase of the line voltage has reversed. Under this circumstance the voltage across the second ignitron may not rise sufficiently to permit firing of that ignitron until some time after the line voltage itself actually switches polarity. If this condition exists, it is possible for the system to misfire since the self-excitation firing system would discharge capacitor 20C or 21C into the ignitor circuit at the appropriate time even through the anode voltage of the associated ignitron may not have risen sufficiently to permit firing, and it is possible for the energy stored in the capacitor to be fully dissipated before the anode voltage rises adequately to permit conductiion in the ignitron. The ignitron anode voltage sensing system obviates this possible malfunctioning.
Upon the closure of the no-weld switch NWCR (FIG. 1) primary winding TSP is connected between the anodes of the two ignitrons 1G1 and IG2 in series with a pair of protective fuses. The voltage across that winding will therefore vary in accordance with the difference between the voltages at the anodes of the two ignitrons. When the anode voltage of either ignitron rises with respect to the other, a voltage is induced across the secondary winding T5S, which is full-wave rectified by rectifiers 7RE and 8RE and applied via resistor 64R to the upper electrode of rectifier l7RE, the lower electrode of which is connected to a suitable source of reference voltage such as positive 12 volts. A rise in voltage at the upper electrode of rectifier 17RE toward the clamped value of 12 volts is communicated via a network comprising resistors 70R and 71R to the base of transistor Q5 which is interconnected with transistor 06 as a form of Schmitt trigger circuit similar to the circuit including transistors Q10 and Q11 previously described. In response to this signal, transistor Q5 abruptly conducts to saturation and transistor O6 is abruptly cut off, transmitting a positive-going pulse to the base of transistor Q7 via a coupling network comprising capacitor l7C and resistors 60R and 42R. As a result, transistor O7 is driven effectively to saturation so that its collector output voltage, applied to conductor 12, falls essentially to ground potential. This voltage is applied to the bases of transistors Q17 and Q19 to enable both of those driving circuits to actuate their associated firing circuit. However, this does not occur until the voltage between the anodes of the two ignitrons has actually changed and been sensed so as to prevent the above-noted misfiring.
In the arrangement illustrated in FIGS. 3 and 4 of the drawings, a pair of back-to-back connected controlled rectifier devices MG and 21G selectively control the energization of the welder transformer T from the source S. The controlled rectifiers MG and 21G are preferably ignitrons although they can be, of course, other devices including appropriate solid state devices. The firing circuits 100 and 102 for the ignitrons 21G and HG are so-called anode firing systems in that each is connected between the anode and the igniter of its respective ignitron. The firing device is a silicon controlled rectifier such as SCR2 in firing circuit (the other firing circuit 102 being identical). A positivegoing input pulse applied via transformer ST and through closed weld switch SWB4 is filtered by a network including resistor R3 and capacitor C2 and applied between the gate and cathode of rectifier SCR2. Diode RE2 connected between the gate and cathode of that rectifier is poled to pass negative-going pulses and to prevent the gate from going negative relative to the cathode. If the anode of rectifier SCR2 is positive relative to the cathode thereof at the instant of the application of the pulse (as other circuits to be described insure), current will flow from source S via line L2, through the primary of welder transformer T15, through rectifier SCR2, through serially interconnected diodes RES and RE6, through choke L3 and resistor R80, the igniter and cathode of ignitron 216, and back to the source via line L1. Rectifiers RES and RE6 are shunted with individual voltage balancing resistors R8 and R9, respectively. Current through the noted path fires ignitron 21G to connect the welder transformer T15 across source S. After ignitron 21G has extinguished, firing circuit 102 fires ignitron 1IG during the next halfcycle to reconnect welder transformer T15 across source S, with the current during successive half-cycles flowing in opposite directions through the primary winding of the welder transformer.
As in the system of FIGS. 1 and 2, if the pulse is applied to the firing for one ignitron before the other ignitron has extinguished (and hence before the anode to cathode voltage of the subject ignitron is correct), the voltage pulse will not produce firing of the ignitron. Since no additional firing pulse can be applied during that same half-cycle, the ignitron will not be actuated during that half-cycle. Accordingly, means are provided to insure that the voltage pulse which is applied to the firing circuit 100 or 102 will not be applied unless the associated ignitron is otherwise prepared to fire.
This is accomplished by sensing the voltage between the anodes of the two ignitrons and hence the voltage across each of the ignitrons, and particularly, sensing the change of that voltage which occurs as a result of the changes of the conductivity of the two ignitrons. The primary winding of a sensing transformer 4T is connected between tha anodes of the ignitrons H6 and 216 and hence is connected across each of those ignitrons. The center tap of the secondary winding of transformer 4T is grounded and the portion of that secondary winding between the center tap and line 104 serves as the voltage supply for transistor 30 in the firing system for ignitron 21G. Correlatively, the other half of that secondary winding serves as the voltage supply for a corresponding transistor of the firing system for ignitron lIG.
As will be seen, transistor 30 is signaled that ignitron 21G should be fired by an abrupt reduction in the voltage applied to its base. Prior to that event, and at the instant (in the assumed stage of functioning) ignitron 11G extinguishes, the voltage across that ignitron abruptly rises to produce a voltage across the secondary winding of transfonner 4T of a polarity such that conductor 14 is positive relative to the grounded center tap. As a result, current flows from conductor 104, rectifier RE39, resistor R48, through the collector and emitter of transistor 30 to ground, a positive voltage being provided to the base at this time, as will be seen. As a result, the potential of the collector of transistor 3Q will drop to a low positive value (approaching ground). When the potential at the base of transistor 3Q is subsequently reduced, which occurs at a time during the subject half-cycle controlled by the heat control circuit, as will be described, transistor 3Q becomes non-conductive and the potential at its collector abruptly rises if but only if transformer 4T is supplying a proper voltage thereto, which occurs only if ignitron lIG has extinguished. If ignitron 1IG has not extinguished at the instant that the voltage at the base of transistor 30 is lowered, the voltage at the collector of transformer 3Q will not rise until that ignitron does extinguish. Until the voltage at the collector of transistor 3Q does rise, no firing pulse can be transmitted to the firing circuit 100, and in this manner it is insured that the-firing pulse will not be applied to ignitron 216 until ignitron 116 has extinguished and the voltage across ignitron 210 is appropriate.
When the voltage at the collector of transistor 30 rises, that voltage isapplied across a circuit including resistor R49, four-layer diode RE52, and resistor R52 (shunted by capacitor C22). While the maximum positive voltage at the collector of transistor 30 is limited by clamping diode Re41, the voltage is adequate break down four-layer diode RE52 and to apply a voltage to the gate of silicon controlled rectifier SCR6.
Silicon controlled rectifier SCR6 is energized from an alternating current source 65 which is preferably the same as or derived from the source S. The phasing is such that the alternating current applied across rectifier SC R6 is in its positive half-cycle when the voltage from source S applied across ignitron 21G is in its negative half-cycle. During the half-cycle in which ignitron llG conducts, currents flows from the secondary winding or transformer T, the center tap of which is grounded, lead 108, resistor R51, rectifier RE44, capacitor C24 and back via ground to the center tap of the secondary winding of transformer 10T. Capacitor C24 becomes charged.
When during the next succeeding half-cycle (the half-cycle in which ignitron 2IG is to be fired), the above-described positive voltage is applied to the gate of rectifier SC R6, that rectifier is rendered conductive and capacitor C24 discharges over acircuit from its upper terminal, conductor 110, primary winding of transformer 3T, conductor I12, anode and cathode of rectifier SC R6, and back to the other terminal of capacitor C24. This discharge current through the primary winding of transformer 3T creates the pulse which fires the firingv device SCR2 to fire ignitron 2IG. It will further be observed since capacitor C24 is charged during one half-cycle and discharges its energy into the firing circuit during the next half-cycle, there is but one opportunity to fire ignitron 2IG in any given hIaf-cycle, illustrating one aspect of the significance of the means for insuring that the ignitron is prepared to fire before the pulse is transmitted.
Corresponding controls and protections are provided for the firing circuit 102 for ignitron lIG, with the requisite phase reversal. Thus, in the firing system for ignitron lIG, the voltage counterpart of that appearing at the collector of transistor 30 is derived from the other half of sensing transformer 4T, that is, from the portion of the secondary winding thereof between conductor I14 and the grounded center tap. Similarly, the voltage for charging the counterpart of cpacitor C24 in the other firing system is derived from the opposite phase of the voltage appearing across transformer 10T.
As noted above, the time at which the potential at the base of transistor 30 is changed determines the firing point of ignitron 2IG. The time in the cycle at which this event occurs is determined by a heat control circuit. That circuit can be of the type above discussed in connection with FIGS. 1 and 2, in which timing is initiated from a preselected instant on the line voltage wave and continues for an interval determined by the charging rate of a capacitor (capacitor 2C in FIG. 2) through a resistance network including a heat adjusting variable resistance (VR2 in FIG. 2) and a power factor adjusting variable resistance (VR3, FIG. 2).
The power factor adjusting variable resistance in FIG. 2 is provided to accommodate the delay between the zero degree (or 180) point on the line voltage waveand the point at which the ignitron anode voltage actually reverses due to the inductive component of the resistance-welder load. Conventional heat controls normally delay the firing of the ignitrons (or other controlled rectifier devices) for a selectable angle measured from a reference point on the line voltage curve, usuallly the zero degree point (and 180 point) on that curve. A 100 percent heat setting would result in the application of igniter current to the ignitrons at that point. However, the ignitron cannot fire there unless the voltage between its anode and the mercury pool is of the correct polarity and of adequate amplitude. With back-to-back connected ignitrons, that condition cannot exist until the other ignitron of the pair has extinguished. A characteristic of controlled rectifiers such as ignitrons is that once fired, they will continue to conduct as long as a minimum holding current is maintained. With the normal inductive load in a resistance welding system, the current lags the voltage, so that the conductive ignitron will normally continue to conduct after the line voltage has reversed and after the zero degree (or 180) point on the line voltage curve. An example is represented in FIG. 5. During a portion of the first negative half-cycle of the line voltage curve E, the negative ignitron is conducting. At the point marked zero degrees on that curve, the line voltage reverses. However, the current through that ignitron and the load lags the line voltage due to the inductive component of the load. In the illustration of FIG. 5, it is assumed that the load is such that the current 1 lags the voltage by about 37 percent power factor). Therefore, the conducting ignitron will continue to conduct after the noted zero-degree point until the current through the ignitron falls below the holding value. When that occurs, the voltage at the positive (nonconducting) ignitron reverses and firing at precisely that point (which is later than the zero degree point on the line voltage wave) would produce percent heat.
In the customary prior practice, the effect of the power factor angle is compensated for by adjusting the power factor variable resistor (such as VR3 in FIG. 2. In common practice, a typical workpiece is inserted during initial setup, the heat control is set at 100 percent, and the power-factor control is then adjusted to equal, approximately, the delay between the zero degree point (or point) of the voltage wave and the point of extinction of the conducting ignitron with that load (which approaches the zero degree point of the current curve I). In many cases,-this adjustment is rarely changed even through different types or thicknesses of workpieces are welded with a resultant shift of the power-factor of the load.
In the system illustrated in FIGS. 3 and 4, power factor adjustment is performed automatically and continually, and the timing of the delay interval (selected by the heat control) is initiated not at the zero degree point (or 180 point) of the line voltage wave or at any other fixed point in reference to the line voltage wave, but rather at the point of voltage reversal across the ignitrons. Thus, timing is initiated at a point which is effectively equal to the power factor angle.
In the representation of FIG. 5, the current has fallen below the holding value at point A, which is close to the zero point of the current I. At the point, the voltage across both ingitrons quite abruptly jumps to the instant line voltage value, that is the value at the point marked B. The voltage across the ignitrons then follows the sinusoidal line voltage curve unit such time as the heat control fires the positive ignitron. In the illustration of FIG. 5, this is assumed to occur at the point C. Upon the firing of the positive ignitron the voltage across both ignitrons abruptly drops to a low positive value determined by the drop across the ignitron. This is illustrated as voltage D in FIG. 5. At the 180 point on the voltage curve E the line voltage reverses. However, since there is still substantial current flow through the conducting ignitron, the voltage across the ignitrons continues at the level D unit the current drops below the holding value, at point F. Thereupon, the conducting positive ignitron extinguishes and the voltage across both ignitrons abruptly changes to the instant negative value of the line voltage curve, as illustrated at point G. The voltage across the ignitrons then follows the sinusoidal line voltage curve until it reaches point H at which time the firing circuit fires the negative ignitron 11G which produces a drop of the voltage across both ignitrons to the level labeled .l on FIG. 5.
In the system of FIGS. 3 and 4, the initiation of the timing of the fire delay in each half-cycle is accomplished by sensing the reversal of the voltage across the ignitrons occuring as at points A, and F, in FIG. 5, and this sensing is accomplished with the same transformer 4T which is utilized to insure against premature application of the firing pulses to the firing circuits. Thus, the two ends of the center-tapped secondary winding of transformer 4T are connected via conductors 114 and 104 to respective ones of two diodes RE9 and RE12. During the period of conduction of either of the ignitrons, when the voltage across the ignitrons (and between their anodes) is continuing at a steady level, no voltage is induced across the secondary winding of transformer 4T. However, at the extinction of the conducting one of the two ignitrons, the voltage across the ignitrons (and the voltage between their anodes) abruptly changes in a positive or negative sense, as at points A and F in FIG. 5. At that instant, a voltage is generated across the secondary winding of transformer 4T and conventional current flows from a positive source of potential through resistor R68, resistor R66, the appropriate one of the rectifiers RE9 and RE12, through conductor 104 or 114 and to that half of the secondary winding of transformer 4T which is at that instant negative. This produces an abrupt reduction in the potential at the base of transistor 4Q and terminates conductivity of that device. As a result, the potential at the collector of transistor 40 rises and this increased positive potential is applied via conductor 122 to establish current flow through resistor R58, through a'heat range adjusting potentiometer P2 (with a portion of the current flowing to ground through resistor R61) through the heat adjust variable resistor P5, resistor R60 and through capacitor C27 to ground. This charging current will continue for the requisite period because following the extinction of the previously conducting ignitron and the resultant abrupt change of voltage across that ignitron, the voltage between the anodes of the two now non-conducting ignitrons follows the sinusoidal line voltage wave to continue to produce a voltage across the secondary winding of transformer 4T to continue to hold transistor 40 nonconductive. The charging of capacitor C27 through the noted resistive path times the delay in the half-cycle before the firing of the ignitron, that is, it times the firing angle. It will be observed that the charging was initiated at the instant of extinction of the previously conducting ignitron so as to achieve automatic power factor adjustment.
When the voltage across capacitor C27 has risen to the preselected value, unijunction UJT3 becomes conductive, in the manner described, producingan increase in the potential at the first base due to the voltage drop across load resistor R59, and this increase in voltage is applied via diode RESO to the gate of silicon controlled rectifier SCR7. That rectifier is accordingly rendered conductive, with current flowing from the secondary winding of transformer 10T, through the full wave rectifier network RE47 and RE48, load resistor R55, through rectifier SCR7 and to ground. As a result, the potential at the anode of rectifier SCR7 drops to a lower positive value approaching ground, and this reduction in potential is communicated via diode RE46, conductor 126 and resistor R46 to the base of transistor 30 to turn that transistor off as previously discussed. This produces the firing pulse to fire the associated ignitron.
Upon that firing, the voltage appearing across the primary winging of transformer 4T drops to the arc value, and during the steady-state condition, no signal voltage is applied to the base of transistor 40 and that base is driven substantially positive by the potential applied thereto through resistor R68. Transistor 4Q thereupon becomes conductive and reduces the potential at the second base of unijunction UJ P3 close to ground value. This permits capacitor C27 to discharge in preparation for the next half-cycle of operation of the system.
While the operation of the system has been described in connection with a typical cycle, it will be understood that various firing angles may be selected and that if desired, and preferably, a minimum delay (such as the 87 V2 delay above-described in connection with FIGS. 1 and 2) may be imposed on the first half cycle of operation of the system during each weld. This may be accomplished in any suitable fashion such as by the use of an and gate similar to that utilized in the system of FIGS. 1 and 2.
While the expression controlled rectifier device" has been used in this specification, it will be appreciated that the ignitrons, or their equivalents, do not truly perform a rectifying function in the systems of the drawings, and the use of that term is not intended to require that the devices in fact rectify current. It will be further recognized that transformer 4T in the system of FIGS. 3 and 4 serves as sensing means for effectively sensing the power factor of the load and produces a signal indicative thereof. The heat control circuits are controlled by the sensing transformer, or equivalent sensing means in or across a portion of the load circuit, by virtue of their responsivity to the signal produced by that transformer. The heat control itself is, of course, basically a form of timing means which times the delay which shall occur between the instant of initiation of the timing and the instant at which the ignitrons or their equivalent are fired in each half-cycle. In the system of FIGS. 3 and 4, that timing is actually initiated by the sensing means. That initation occurs, in the illustrated arrangement, at the instant that the conducting ignitron becomes non-conductive, which coincides, for all practical'purposes, with the point at which the lagging currentreaches Zero. Hence, the signal produced by the transformer is indicative of the delay between the zero point on the voltage wave and the zero point on the current wave and hence is indicative of the power factor of the load.
While the arrangement of FIG. 4 utilizes ignitrons controlled by silicon control rectifiers, it will be recognized that in accordance with convention practice the SCRs may be used directly as the controllable means for interconnecting the power source and the welder transformer (as by omitting the ignitrons in FIG. 4, connecting the cathode of silicon controlled rectifier SC R2 directly to the anode of the corresponding silicon controlled rectifier in firing circuit 102 and connecting the cathode of the latter directly to the anode of the former and omitting the ignitrons) if the load requirements are such that they can be handled, in a given installation, by the silicon controlled rectifiers.
While the above described embodiments provide a satisfactory control of the energy being fed to a weld in response to variations in power factor, it has been found that a fully automatic full range system is desired under many circumstances.
Referring to FIG. 5, it is seen that the reversal of the voltage and achieving line voltage across the ignitrons, as sensed by the transformer interconnected across the ignitrons, is approximately equal to the lag or zero current point of the current wave form due to the inductive character of the load. Thus, it is seen that the jump in voltage from approximately zero volts across the ignitrons due to the conduction of one ignitron at point A to the line voltage at point B occurs approximately at the zero current cross-over point. Thus, the relationship between the zero current cross-over point and the zero voltage cross-over point may be termed the power factor angle.
The remaining portion of the voltage cycle is then seen to be 100 percent for the particular load, correlated with its inductive characteristics, connected to the power supply. It is the energy under this remaining curve which is controlled to provide from zero to approximately 100 percent of this energy to the load, also known as the percent heat.
In accordance with the system of the embodiment to be described in conjunction with FIGS. 6-11, it has been found that an accurate control of the heat being supplied to the load may be achieved from zero to approximately 98 percent heat through a system which includes generating a straight line wave form which starts at the zero voltage point or the zero cross-over point of the voltage and advances along a preselected line with a preselected slope until such time as the reversal of the voltage across the ignitrons or the achieving of line voltage is sensed. At this time, the system of the present invention generates a second straight line having a slope which is selected in accordance with the percent heat desired at the load, this percent heat being apercentage of the energy remaining under the remainder of the curve to the zero voltage cross-over point. The slope of this latter straight line is increased toward infinity to achieve very nearly full (100 percent) heat or is decreased to a point which may coincide with the original slope to achieve very nearly zero heat.
Further, a reference level is generated within the system to be described, which reference level is compared with the magnitude of the signal generated after the achieving of the zero current cross-over point. When the magnitude of the signal represented by the second line equals or slightly exceeds the reference level, the ignitrons are fired to supply heating energy to the load.
With the system described above, the control of the present heat is a variant which is independent of the power factor of the load and, for a particular setting of the slope of the second line signal, will generate a preselected percent heat relative to the total energy under the envelope formed by the line voltage signal after the zero cross-over point. Thus, for a preselected slope of the second line, taking for example percent heat, the second signal will equal or slightly exceed the reference level at a point which permits three-fourths of the energy level remaining under the curve after firing to be fed to the load circuit.
Referring now to FIGS. 6 through 11, it has been found that completely automatic phase shift correction can be achieved by sensing when the current drops to zero relative to the point in time when the voltage drops to zero. With the system of the present invention, a circuit is provided which generates a linearly increasing straight line curve having a preselected slope. This curve is generated at the point in time when the voltage is at a zero potential.
When the system senses the current crossing over at the zero point, a second timer is switched into circuit with the timing system, the second circuit generating a linearly increasing signal having a straight line configuration, the slope of this latter signal being selectively variable to select the percent heat which is to be fed to the work. In this particular situation, it has been found that a single slope setting, as for example for 75 percent heat, will provide the load with 75 percent of the energy available from the zero current cross-over point to the end of the voltage half-wave. When the second curve reaches a preselected pulsing level, the ignitrons are fired to cause energy to flow through the electrodes to the workpiece. As is readily apparent, at a percent phase shift, the current will not cross the zero point prior to the end of the voltage half-wave and the second curve will not be generated for any cycle in which this occurs. This is due to the fact that there is no energy remaining between the zero current crossover point and the end of the voltage half-wave.
Referring now to FIG. 6, there is illustrated, in the upper portion thereof, a voltage-versus-time curve 100